Bull***t...
Metrum Acoustics and Soekris are just two examples of R2R dac architecture currently in production, and they have an excellent reputation for sound quality.
Those are not commercially available DAC ICs. I really don't care what audiophile boxes of junk with poor measurements are out there.
Same thing with the Schiit boxes using the AD non-audio DAC that measures like crap. You are free to believe in the false narrative, hype, and myths of course.
Recommended Components: 2018 Edition Digital Processors | Stereophile.com
A+ list seems to be totally dominated by delta sigma converters.
Last edited:
Transient dac ICs are OEM available, just like Soekris R2R technology.
Can you back up your statement with proper evidence?
Those are not integrated circuits, those are PCBAs. I have said all that is needed on this.
You guys remind me something I wanted to ask for a long time..
Many people here in DIYAudio mention that one of the biggest disadvantages of Voltage-Out DAC Chips, is that since the I/V part (and other stuff) are built-in the chip, it is hard (or impossible) to disable them.
On the other hand, old MultiBit DAC chips (that don't include all the integrated functionality) are not being manufactured anymore... and so are very hard to get.
So my question is:
Can simply using many resistors (R-2R) to create a DAC, do the job well? (for audio)
Is that what the 2 companies daanve mentioned do?
In this way, every one here can PCB Print a small board, and then solder the resistors to it..
It will not be compact like a small chip,
but for people who need it for home use, and not mobile/portable, the size of it does not matter, only the sound quality that is gotten from it..
What do you think?
Last edited:
The OP question was obviously about chips. Please, just stop with this audiophile module crap. Every other time over the last decade when this question comes up it is always in the context of chips, because the OEM module should already include an I/V converter if one is necessary.
Last edited:
It's much more complicated than that if you want it to perform well at all.
Resistor ladder - Wikipedia
You can see from the measurements of the Soekris on ASR that it is mostly unable to match PCM1702 in all respects which was discontinued over 10 years ago and released in 1995.
Metrum, I have only seen one set of measurements for one of their products and they were very poor.
I am not bashing Soren / Soekris. He's done good work on that and it is not an easy task. I just don't see the point in it, personally.
Last edited:
Chris and daanve,
please don't argue,
the info you give is useful.
Chris:
my question was about chips, but I am not limiting myself to chips.
If PCB Printing a small module can give a better sound, I am willing to do it.
PCB Printing today is accessible even in small quantities,
same for PCB Assembly, for those who need it.
Apparently it would be a very important thing not because of the ruled out capacitor , but because it allows the output to be higher without hitting the rail supplies.
By the way, with the right input, this dac sounds very well as one of my friends reported after using an x-mos i2s-asyncronous usb module.
Unfortunately, nowadays v-out dacs don't have the option of having external capacitance of the i/v stage so it all relies on two aspects:
1.the i/v stage linearity when close to rail supplies
2.max rail supply voltage.
There's a marketing pressure for using USB supplied audio equipment which makes it for the fact that they need to make better rro's .
There's also an even bigger problem, pending from the beginnings of dac technology .
The standard output of the digital players was established at 2vRMS which led to very big problems with v-out dac's.
You need more than +-5v supply rails for the v-out dacs to have similar linearity with the i-out dac plus separate op-amps.
The output is already very high for most of the audio amplifiers taking the signal and the active low pass filters after the dac needed to work at unity gain and being also very stable with very high frequency input noises so most of the op-amps used behave poorly.If they only made it 0.5...1v output, the i/v stage would be able to drive 4x lighter passive filters or the cheapest op-amps with a 2x...4x gain could make the best output, compensating with their better CMRR and PSRR factors due to higher gains for the lower SNR of a lower gain I/V stage.
I found out that the old days v-out dacs that allowed external Rfb and Cfb could work way more linear if i made the output of the i/v stage smaller, 1vRMS instead of 2vRMS so that the output op-amp will never clip.
The only way the manufacturers can make better v-out dacs with the market pressure is to make the digital conversion less noisier and lower the output in case they can't make better rail to rail op-amps.It seems though that they already did good enough v-out dac's but they also need to sell very expensive op-amps for the audiophile market.
I personally think that they already have the technology to deliver the perfect v-out dac but they are also the same manufacturers of the expensive op-amps...
There are lots of cheap technologies that could be used in our favor but they need to sell 5...10 pounds op-amps each...
Last edited:
Just look at the Soekris DAM thread. The resistor array is only one component of the converter. The rest is done in the FPGA. It's not easy at all. If it were, you would see more of them. Even with all that work, the module doesn't match the performance of an $8 DAC chip from AKM.
Some people might say that they think it sounds better. That's pure opinion. There are hugely expensive products with effusive reviews that use off-the-shelf DAC chips, too.
Again I'm not trying to bash the Soekris stuff. Soren has put a lot of work into those and knows what he is doing. They have excellent SNR, but the distortion specs are not that great. I just don't see the need to avoid sigma delta.
Last edited:
Thank you dreamth for the explanation..
Too bad then..
BTW,
no-one here said any opinion regarding the DAC product I mentioned in post #10 in this thread..
I see.
Too bad then..
BTW,
no-one here said any opinion regarding the DAC product I mentioned in post #10 in this thread..
Last edited:
If you want the very best of the voltage out chips, then look for a design based on AK4493, AK4497, or CS43198. PCM5102 probably sounds just fine, but it's a part designed for design convenience.
That's not entirely true; absurdly-high-dynamic-range audio DACs are usually multibit sigma-delta designs, but there certainly are single-bit audio DACs in quantity production (in IC form).
They might exist, maybe integrated in codecs, but I haven't seen a datasheet that explicitly says that in a long time. There might be something in production but I thought most designs moved to a few bits even in the low-end of the market.
http://www.cscamm.umd.edu/programs/ocq05/adams/adams_ocq05.pdf
The quality of the USB interface and the power supply (USB as well with likely a switcher) could be more important than the quality of the dac chip.
I would suppose that SACD player application DAC ICs are single-bit delta-sigma, however, I don't know that for certain.
You can do a switching power supply that performs just as well as a linear supply, but yeah in a $35 eBay DAC I would be skeptical of the quality of implementation.
I think the SA9023 is fine but it depends if it’s being used in asynchronous mode or not.
The newer ones aren’t, even DSD1792 is internally identical to PCM1792. I’m sure there’s some old part that Sony used or something floating around though.
Just another reason DSD was ill conceived. I find it funny that most recordings convert to / from DXD (PCM) and then the playback system is often doing conversion back.
Last edited:
I referred to the FX-01 USB dac (post #10) which has a budget chip (PCM5102).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.